feat: prepare Radixor 3.0.0 with contracted tries and compiled patch commands
Introduce contracted compiled patch tries for faster lookup, make compiled patch commands the primary runtime path, refresh stemmer benchmarks and documentation, and restructure the documentation for 3.0.0 onboarding. BREAKING CHANGE: Radixor 3.0.0 promotes compiled patch-command APIs and new compiled trie artifacts as the primary runtime integration model.
This commit is contained in:
223
gradle/lucene-benchmarks.gradle
Normal file
223
gradle/lucene-benchmarks.gradle
Normal file
@@ -0,0 +1,223 @@
|
||||
import org.gradle.plugins.ide.eclipse.model.SourceFolder
|
||||
|
||||
|
||||
def luceneVersion = '10.5.0'
|
||||
def luceneRootRelativePath = 'third-party/lucene'
|
||||
def luceneSourceArtifacts = ['lucene-analysis-common', 'lucene-analyzers-common']
|
||||
def luceneSourceDirectory = layout.buildDirectory.dir("${luceneRootRelativePath}/source/analyzers-common")
|
||||
|
||||
def luceneGeneratedSourceDirectory = layout.buildDirectory.dir('generated/sources/lucene')
|
||||
def luceneGeneratedPorterFile = luceneGeneratedSourceDirectory.map { it.file('org/egothor/stemmer/benchmark/LucenePorterStemmerCopied.java') }
|
||||
|
||||
def luceneSourceDownloadFile = layout.buildDirectory.file("${luceneRootRelativePath}/lucene-${luceneVersion}-sources.jar")
|
||||
|
||||
dependencies {
|
||||
jmhImplementation "org.apache.lucene:lucene-analysis-common:${luceneVersion}"
|
||||
jmhImplementation "org.apache.lucene:lucene-analysis-stempel:${luceneVersion}"
|
||||
jmhImplementation "org.apache.lucene:lucene-analysis-morfologik:${luceneVersion}"
|
||||
}
|
||||
|
||||
def buildLuceneSourcesName = { final String artifact ->
|
||||
"${artifact}-${luceneVersion}-sources.jar"
|
||||
}
|
||||
def buildLuceneSourcesUrl = { final String artifact ->
|
||||
"https://repo1.maven.org/maven2/org/apache/lucene/${artifact}/${luceneVersion}/${buildLuceneSourcesName(artifact)}"
|
||||
}
|
||||
|
||||
def isLuceneSourcesDownloadable = { final String artifact ->
|
||||
try {
|
||||
final URL sourceUrl = new URL(buildLuceneSourcesUrl(artifact))
|
||||
final java.net.HttpURLConnection connection = (java.net.HttpURLConnection) sourceUrl.openConnection()
|
||||
connection.requestMethod = 'HEAD'
|
||||
connection.instanceFollowRedirects = true
|
||||
connection.connectTimeout = 10000
|
||||
connection.readTimeout = 10000
|
||||
final int responseCode = connection.responseCode
|
||||
connection.disconnect()
|
||||
return responseCode == 200
|
||||
} catch (Exception ignored) {
|
||||
return false
|
||||
}
|
||||
}
|
||||
|
||||
def downloadLuceneSourcesJar = { ->
|
||||
final File targetFile = luceneSourceDownloadFile.get().asFile
|
||||
|
||||
for (String artifact : luceneSourceArtifacts) {
|
||||
if (!isLuceneSourcesDownloadable(artifact)) {
|
||||
continue
|
||||
}
|
||||
|
||||
final String sourceUrl = buildLuceneSourcesUrl(artifact)
|
||||
final File tempFile = new File(targetFile.parentFile, "${artifact}.${luceneVersion}.tmp")
|
||||
try {
|
||||
new URL(sourceUrl).withInputStream { inputStream ->
|
||||
tempFile.parentFile.mkdirs()
|
||||
tempFile.withOutputStream { outputStream ->
|
||||
outputStream << inputStream
|
||||
}
|
||||
}
|
||||
if (!tempFile.exists() || tempFile.length() <= 0L) {
|
||||
throw new GradleException("Downloaded Lucene source artifact for ${artifact} was empty.")
|
||||
}
|
||||
targetFile.delete()
|
||||
if (!tempFile.renameTo(targetFile)) {
|
||||
throw new GradleException("Failed to persist downloaded Lucene source artifact for ${artifact}.")
|
||||
}
|
||||
return
|
||||
} catch (Exception ignored) {
|
||||
tempFile.delete()
|
||||
}
|
||||
}
|
||||
|
||||
throw new GradleException(
|
||||
"Failed to download Apache Lucene source artifacts ${luceneSourceArtifacts} for version ${luceneVersion}.")
|
||||
}
|
||||
|
||||
def luceneSourceClasspathPath = provider {
|
||||
project.relativePath(luceneGeneratedSourceDirectory.get().asFile)
|
||||
}
|
||||
def luceneEclipseClasspathAttributes = [
|
||||
gradle_scope : 'jmh',
|
||||
gradle_used_by_scope: 'jmh',
|
||||
test : 'true'
|
||||
]
|
||||
|
||||
def isAbsoluteClasspathPath = { String path ->
|
||||
path.startsWith('/') || path ==~ /^[A-Za-z]:[\\\/].*/
|
||||
}
|
||||
|
||||
def luceneGeneratedPorterNotice = '''
|
||||
/**
|
||||
* Generated at benchmark execution time from Apache Lucene source.
|
||||
*
|
||||
* This source copy is compiled only for the JMH benchmark source set and is
|
||||
* not committed as production code.
|
||||
*/
|
||||
'''
|
||||
|
||||
def transformPorterStemmerSource = { final File sourceFile, final File targetFile ->
|
||||
if (!sourceFile.exists()) {
|
||||
throw new GradleException("Apache Lucene PorterStemmer source was not available at ${sourceFile}.")
|
||||
}
|
||||
|
||||
final String sourceText = sourceFile.getText('UTF-8')
|
||||
String transformedText = sourceText
|
||||
|
||||
if (transformedText.contains('package org.apache.lucene.analysis.en;')) {
|
||||
transformedText = transformedText.replaceFirst(/(?m)^\s*package\s+org\.apache\.lucene\.analysis\.en\s*;/,
|
||||
'package org.egothor.stemmer.benchmark;')
|
||||
} else {
|
||||
throw new GradleException(
|
||||
'Expected Lucene package-private PorterStemmer in org.apache.lucene.analysis.en package was not found in downloaded source.')
|
||||
}
|
||||
|
||||
transformedText = transformedText.replaceFirst(/(?m)^\s*class\s+PorterStemmer\s*\{/, 'public final class LucenePorterStemmerCopied {')
|
||||
transformedText = transformedText.replaceFirst(/(?m)^\s*public\s+PorterStemmer\(\)/, 'public LucenePorterStemmerCopied()')
|
||||
|
||||
if (!transformedText.contains('class LucenePorterStemmerCopied')) {
|
||||
throw new GradleException("Failed to rename PorterStemmer class when generating ${targetFile}.")
|
||||
}
|
||||
|
||||
targetFile.parentFile.mkdirs()
|
||||
targetFile.text = transformedText
|
||||
}
|
||||
|
||||
def resolveLucenePorterStemmerSource = { ->
|
||||
final File sourceRoot = luceneSourceDirectory.get().asFile
|
||||
final List<String> candidates = [
|
||||
'org/apache/lucene/analysis/en/org/apache/lucene/analysis/en/PorterStemmer.java',
|
||||
'org/apache/lucene/analysis/en/PorterStemmer.java',
|
||||
'org/apache/lucene/analysis/en/org/tartarus/snowball/ext/PorterStemmer.java'
|
||||
]
|
||||
|
||||
for (String candidate : candidates) {
|
||||
final File file = new File(sourceRoot, candidate)
|
||||
if (file.exists()) {
|
||||
return file
|
||||
}
|
||||
}
|
||||
|
||||
final FileTree porterCandidates = fileTree(sourceRoot).matching { include '**/PorterStemmer.java' }
|
||||
for (File file : porterCandidates.files) {
|
||||
if (file.text.contains('class PorterStemmer') && file.text.contains('package org.apache.lucene.analysis.en;')) {
|
||||
return file
|
||||
}
|
||||
}
|
||||
|
||||
throw new GradleException('Unable to resolve Lucene PorterStemmer source file from extracted artifact.')
|
||||
}
|
||||
|
||||
tasks.register('downloadLuceneAnalyzersSources') {
|
||||
group = 'build setup'
|
||||
description = 'Downloads Apache Lucene analysis sources for benchmark-only code generation.'
|
||||
|
||||
outputs.file(luceneSourceDownloadFile)
|
||||
|
||||
doLast {
|
||||
if (!luceneSourceDownloadFile.get().asFile.exists()) {
|
||||
downloadLuceneSourcesJar()
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
tasks.register('extractLuceneAnalyzersSources', Copy) {
|
||||
group = 'build setup'
|
||||
description = 'Extracts Apache Lucene analysis source JAR for benchmark-only extraction.'
|
||||
|
||||
dependsOn(tasks.named('downloadLuceneAnalyzersSources'))
|
||||
|
||||
from(zipTree(luceneSourceDownloadFile))
|
||||
into(luceneSourceDirectory)
|
||||
}
|
||||
|
||||
tasks.register('generateLucenePorterStemmerCopied') {
|
||||
group = 'build setup'
|
||||
description = 'Generates LucenePorterStemmerCopied into the build-only benchmark source directory.'
|
||||
|
||||
dependsOn(tasks.named('extractLuceneAnalyzersSources'))
|
||||
|
||||
inputs.dir(luceneSourceDirectory)
|
||||
outputs.file(luceneGeneratedPorterFile)
|
||||
|
||||
doLast {
|
||||
final File sourceFile = resolveLucenePorterStemmerSource()
|
||||
transformPorterStemmerSource(sourceFile, luceneGeneratedPorterFile.get().asFile)
|
||||
}
|
||||
}
|
||||
|
||||
sourceSets {
|
||||
jmh {
|
||||
java {
|
||||
srcDir(luceneGeneratedSourceDirectory)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
tasks.named('compileJmhJava') {
|
||||
dependsOn(tasks.named('generateLucenePorterStemmerCopied'))
|
||||
}
|
||||
|
||||
eclipse {
|
||||
classpath {
|
||||
file {
|
||||
whenMerged { classpath ->
|
||||
String generatedPath = luceneSourceClasspathPath.get()
|
||||
|
||||
classpath.entries.removeAll { entry ->
|
||||
entry.hasProperty('path') && (
|
||||
entry.path == generatedPath ||
|
||||
isAbsoluteClasspathPath(entry.path)
|
||||
)
|
||||
}
|
||||
|
||||
SourceFolder luceneEntry = new SourceFolder(generatedPath, null)
|
||||
luceneEntry.output = 'bin/jmh'
|
||||
luceneEclipseClasspathAttributes.each { String name, String value ->
|
||||
luceneEntry.entryAttributes[name] = value
|
||||
}
|
||||
classpath.entries.add(luceneEntry)
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
Reference in New Issue
Block a user